I would disagree with this. There are African villages where lots of kids die of diarrhea, and when researchers introduced solar water disinfection (essentially put water in a plastic jug and put it in the sun for a while), people wouldn't do it because it signaled that they were low class, despite that fact that lots of child deaths could be prevented.

Similarly, economic returns vs mortality risks of running in a gang.

Similarly, drug addicts and alcoholics.

And don't forget fatties.

Now, one can respond "revealed preferences" and kind of defeat the purpose of calling actions rational or irrational, but people's actions often are not too closely linked to survival.

Well, historically in case of basic subsistence activites, winning meant surviving, and loosing meant dying a horrible death. There are likely some strong adaptations in play here.